• Picks
  • Discover
  • Alternatives
  • About

© 2026 Willet

  • Newsletter
  • About
  • FAQ
  • Terms
  • Privacy & Cookies
Join the Discord (opens in a new tab)
VXControl avatar

PentAGI

Self-hosted AI agents that run autonomous penetration tests

  • Security testing and OSINT
  • Agent frameworks and tools
View on GitHub
Website

What it is

PentAGI coordinates a group of specialized AI agents that plan and perform penetration testing work, running scanners and exploit kits such as sqlmap, nmap, and Metasploit within a sealed-off Docker container. It stores findings in PostgreSQL with pgvector and an optional Neo4j knowledge graph, works with cloud or local LLM providers, and exposes REST and GraphQL APIs. Security engineers and red teams deploy it on servers they control to conduct penetration tests they are authorized to perform.

Images

Facts

Stars
31,483 GitHub stars.
31,483

Forks
5,039
Commits
952
Contributors
15
Language
Go
License
MIT
Last commit
Oct 9, 2026
On GitHub since
January 2025
First release
January 2025
Platforms
web, macOS, Windows, Linux, self-hosted
Self-hostable
Yes
Tags
  • multi-agent
  • agent-sandbox
  • penetration-testing
  • ai-agents
  • vulnerability-scanning
  • knowledge-graph

Last synced from GitHub on Oct 20, 2026

The 10 best overlooked open-source projects, every Monday.

Share and more

Comments

No comments yet. Start the conversation.

Similar projects

  • AIPentest avatar

    CyberStrikeAI

    AI agents for authorized penetration testing and security operations

    7.9k
  • Strix avatar

    Strix

    AI agents that pentest your app and confirm bugs with working exploits

    70.3k
  • Keygraph avatar

    Shannon

    Finds and proves web app and API vulnerabilities with real exploits

    49.1k

What people are saying

Updated Oct 20, 2026
  • Bluesky“PentAGI is a fully autonomous AI red-team tool released as open source. It uses a team of specialized AI agents for research, development, penetration…”Oct 15, 2026 · 6 likes · 1 reply (opens Bluesky in a new tab)
  • Bluesky“vxcontrol/pentagi: Fully autonomous AI Agents system capable of performing complex penetration testing tasks #devopsish”Mar 28, 2026 · 3 likes · 0 replies (opens Bluesky in a new tab)
  • Hacker NewsPentagi: Autonomous AI Agents for complex penetration testing tasksFeb 22, 2026 · 2 points · 0 comments (opens Hacker News in a new tab)
  • Bluesky“GitHub - vxcontrol/pentagi: Fully autonomous AI Agents system capable of performing complex penetration testing tasks · GitHub”Apr 7, 2026 · 1 like · 0 replies (opens Bluesky in a new tab)
  • Bluesky“Fully autonomous AI penetration testing system that runs 20+ security tools in sandboxed Docker environments - like having a senior pentester that never…”Mar 24, 2026 · 1 like · 0 replies (opens Bluesky in a new tab)
  • Bluesky“vxcontrol/pentagi Fully autonomous AI Agents system capable of performing complex penetration testing tasks PentAGI P enetration testing A rtificial G eneral…”Mar 24, 2026 · 0 likes · 0 replies (opens Bluesky in a new tab)